Ensemble Subsurface Modeling With External Models for Lower Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current reservoir modeling methods produce characterization models with significant uncertainty, impacting the accuracy of fluid flow simulation and hydrocarbon production planning.
Innovation Solution
Employ ensemble machine learning prediction, combining multiple machine learning models with data from external models like Kriging to enhance predictive performance and reduce uncertainty, using decision trees and random forest learning methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional reservoir modeling methods are used, then the modeling process can be completed with conventional techniques, but the produced characterization models have significant uncertainty that impacts accuracy
Solution Approach 1:
The patent combines multiple machine learning models (ensemble learning) including decision trees, random forests, and external models like Kriging to create a unified predictive system. This merging of multiple modeling approaches reduces uncertainty and improves the reliability of reservoir characterization models compared to using any single conventional method.
Solution Approach 2:
The invention creates a composite modeling approach by integrating heterogeneous data sources and algorithmic methods (seismic data, well log data, core data, and multiple ML algorithms) into a unified ensemble system. This composite structure leverages the strengths of different models to produce more accurate and reliable predictions than individual components alone.
2Measurement precision
If ensemble machine learning with multiple external models is employed, then predictive performance and model accuracy are enhanced, but the computational complexity and processing requirements increase
Solution Approach 1:
The patent segments the complex ensemble modeling process into distinct modules: data preprocessing, individual model training (seismic, well log, core), external model integration (Kriging), and ensemble aggregation. This segmentation allows each component to be optimized independently while managing overall system complexity through modular architecture.
Solution Approach 2:
The invention introduces intermediary layers including data normalization protocols, feature selection mechanisms, and weighted aggregation functions that mediate between raw inputs and final predictions. These intermediaries simplify the integration of multiple complex models by standardizing data formats and prediction outputs before combining them in the ensemble.
Data Source
AI summary
Method and systems are provided that create one or more models of a subsurface geological formation (such as a reservoir characterization model of a hydrocarbon reservoir or a model of some other subsurface geological formation). The method and systems are configured to extend a machine learning ensemble (such as an ensemble tree-based machine learning model such as a random forest learning model) to use or embed data derived from one or more secondary models as part of the training operations of the machine learning ensemble and online use of the trained machine learning ensemble. Such data can provide information that supplements the information contained in the training data/input data.